CN106326996B - User load prediction method based on electric quantity information - Google Patents
User load prediction method based on electric quantity information Download PDFInfo
- Publication number
- CN106326996B CN106326996B CN201510335111.2A CN201510335111A CN106326996B CN 106326996 B CN106326996 B CN 106326996B CN 201510335111 A CN201510335111 A CN 201510335111A CN 106326996 B CN106326996 B CN 106326996B
- Authority
- CN
- China
- Prior art keywords
- user
- value
- load
- time
- probability
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 40
- 239000011159 matrix material Substances 0.000 claims abstract description 48
- 238000005070 sampling Methods 0.000 claims abstract description 44
- 230000002068 genetic effect Effects 0.000 claims abstract description 9
- 210000000349 chromosome Anatomy 0.000 claims description 38
- 230000005611 electricity Effects 0.000 claims description 16
- 239000003550 marker Substances 0.000 claims description 12
- 108090000623 proteins and genes Proteins 0.000 claims description 7
- 238000005457 optimization Methods 0.000 claims description 2
- 238000010276 construction Methods 0.000 abstract description 8
- 230000035772 mutation Effects 0.000 description 6
- 230000003203 everyday effect Effects 0.000 description 2
- 240000007594 Oryza sativa Species 0.000 description 1
- 235000007164 Oryza sativa Nutrition 0.000 description 1
- 238000004378 air conditioning Methods 0.000 description 1
- 238000010411 cooking Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007717 exclusion Effects 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 235000009566 rice Nutrition 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Abstract
The invention provides a user load prediction method based on electric quantity information, which comprises the following steps of 1: determining a power consumption set Q at a peak load period of a user; step 2: setting the time sequence number i of the user in the peak load period to be 1; and step 3: will use a probability typical value matrix PBi‑1Assignment to matrix of use probability typical values PBiPerforming the following steps; and 4, step 4: constructing a load curve sequence value set Pi(ii) a And 5: optimized use of probability typical value matrix PB by genetic algorithmi(ii) a Step 6: setting i to i + 1; and 7: according to optimized use probability typical value matrix PBiAnd the quantity of various types of household appliances owned by the user, and calculating and predicting the load value of the user by using a Monte Carlo non-sequential random sampling method. Compared with the prior art, the user load prediction method based on the electric quantity information is convenient and quick, improves the prediction accuracy, lays a foundation for reasonable configuration of the capacity of the distribution transformer in the distribution area, and improves the economic operation level and the construction efficiency of the distribution area.
Description
Technical Field
The invention relates to the field of power systems, in particular to a user load prediction method based on electric quantity information.
Background
With the rapid and stable development of economic society, the orderly development of novel urbanization and beautiful country construction, especially with the continuous increase of household appliances of rural residents, the power load of the residents in China continuously increases, and the characteristics of the power consumption of the residents in life are continuously changed. The method can accurately master the user load, particularly the increase condition and the power utilization characteristic of the user load in peak load period, and is the basis for reasonably planning the distribution station and optimizing the operation control of the distribution station. At present, the user power load is generally estimated according to experience in the planning and construction process of an actual power distribution station, the accuracy is not high, and the situation that the power distribution station is heavily overloaded shortly after capacity expansion is carried out in the power distribution station is caused to occur frequently. Therefore, accurate simulation and prediction of user load changes are urgently needed to be solved. At present, a power distribution station only accurately acquires power consumption information for users, electric quantity of peak time periods, flat sections and valley sections of each day can be collected through a user intelligent electric meter, and rated power and owned quantity of household appliances of the users can be acquired through a statistical method.
The method can accurately master the user load, particularly the increase condition and the power utilization characteristic of the user load in peak load period, and is the basis for reasonably planning the distribution station and optimizing the operation control of the distribution station. At present, the user power load is generally estimated according to experience in the planning and construction process of an actual power distribution station, the accuracy is not high, and the situation that the power distribution station is heavily overloaded shortly after capacity expansion is carried out in the power distribution station is caused to occur frequently. Therefore, a user load prediction method for conveniently, quickly and accurately simulating and predicting user load change is needed to lay a foundation for reasonable configuration of the capacity of the distribution transformer in the distribution area and improve the economic operation level and the construction efficiency of the distribution area.
Disclosure of Invention
The invention provides a user load prediction method based on electric quantity information, aiming at accurately mastering the user electric load in the planning and construction process of a power distribution station area.
The technical scheme of the invention is as follows:
the method comprises the following steps:
step 1: obtaining a category set DQ, a power set PDQ, a quantity set NDQ and a use probability initial typical value matrix PB of the household appliances of the user0The time set T of the peak load period and the electricity consumption set Q of the user in the peak load period;
step 2: setting the time sequence number i of the user in the peak load time period to be 1, wherein i is more than or equal to 1 and is less than or equal to m, and m is the total time of the peak load time period;
and step 3: will be provided withUsing a probabilistic typical value matrix PBi-1Assignment to matrix of use probability typical values PBiPerforming the following steps; the PB isi-1Is ti-1Probability typical value matrix, PB, for use by a domestic appliance at a timeiIs tiThe household appliances at the moment use the probability typical value matrix;
and 4, step 4: according to the typical value matrix PB of the use probabilityiCalculating the load value of each moment in the peak load period of the user, and constructing a load curve sequence value set Pi＝{p1,p2,...,pi,...,pm}，piIs tiThe load value of the household appliance at the moment;
and 5: optimizing said usage probability typical value matrix PB by a genetic algorithmi；
Step 6: setting a time sequence number i ═ i +1 in a peak load period;
and 7: if i is less than or equal to m, returning to the step 3; if i is larger than m, the typical value matrix PB of the optimized household appliance use probability is usediAnd the quantity of various types of household appliances owned by the user, and the load value of the user is calculated and predicted by using a Monte Carlo non-sequential random sampling method.
Preferably, the set of categories DQ of the household appliances in the step 1 is { DQ ═ d1,dq2,...,dqj,...,dqnN is the total number of types of household appliances, dqjName of the j type household appliance; the power set PDQ of the household appliance is { PDQ ═1,pdq2,...,pdqj,...,pdqn}，pdqjThe average power of the j type household appliance; the number set NDQ of the household appliances is { NDQ ═1,ndq2,...,ndqj,...,ndqn}，ndqjThe number of the j type household appliances; the set of times of the peak load period T ═ T1,t2,...,ti,...,tm}，tmIs a time value of 24 system time, t is more than or equal to 1m≤24；
An initial typical value matrix of the usage probability of the household appliance
The above-mentioned
Is tiTime j type household appliance dqjThe initial value of the typical value of the use probability;
preferably, in step 1, the set of user power amounts Q ═ Q1,q2,...,qk,...,qd}，qkD is the total number of days for collecting the electricity consumption by the intelligent ammeter;
in the step 4, the Monte Carlo non-sequential random sampling method is used for calculating tiLoad value p of the domestic appliance at any timei；
Preferably, said step 5 optimizes t by genetic algorithmiTypical value matrix PB of use probabilities of time-of-day household appliancesiThe method comprises the following steps:
step 5-1: setting the initial value of the iteration times h as 0, the maximum value of the iteration times h as maxnum, and the number of population scales as ZQnum; will tiTypical value matrix PB of use probabilities of time-of-day household appliancesiEach using the probability to perform corresponding chromosome coding to generate an initial population ZQ0；
Step 5-2: setting the iteration number h as h + 1;
step 5-3: the population ZQh-1Assigns data to a population ZQh；ZQh-1For the h-1 th iteration generated population, ZQhThe population generated for the h iteration; setting population ZQhThe initial value of the middle chromosome marker bit l is 0;
step 5-4: setting the chromosome marker bit l ═ l + 1;
step 5-5: the l chromosome is subjected to inverse coding to obtain tiThe use probability of each type of household appliance at any moment;
and 5-6: according to the said tiThe using probability of each type of household electrical appliance at the moment is obtained by a Monte Carlo non-sequential random sampling methodiMoment of time the load of the household applianceCharge value piAnd updating the load curve sequence value set Pi＝{p1,p2,...,pi,...,pm}；
And 5-7: calculating the load average value Pav of the household appliances of the user in the peak load period according to the updated load curve sequence set P,
and 5-8: acquiring an average load set PG (PG) of peak load periods of a user according to the electricity consumption set Q1,pg2,...,pgk,...,pgd}；pgkD is the total number of days for collecting the electricity consumption of the user by the intelligent ammeter;
and 5-9: the fitness function value epsilon of the l-th chromosome is calculated,
step 5-10: if the chromosome marker bit l is less than ZQnum, returning to the step 5-4; if the chromosome marker bit l is more than or equal to ZQnum, executing the step 5-11;
step 5-11: arranging the chromosomes from small to large according to the numerical value of the fitness function value epsilon, selecting 1 st to ZQnum chromosomes and then executing the step 5-12;
and 5-12: for population ZQhCarrying out whole cross operation of the parent and the gemini single-point genes;
step 5-13: carrying out variation on chromosomes and updating the ZQ population according to the varied chromosomesh；
And 5-14: if the iteration times h are less than maxnum, returning to the step 5-2; if the iteration times h is larger than or equal to maxnum, the chromosome with the optimal fitness function value epsilon is subjected to inverse coding to obtain tiThe using probability optimization value of the household appliance at any moment;
preferably, in step 7, the calculating t by the Monte Carlo non-sequential random sampling methodiLoad value p of the domestic appliance at any timeiThe method comprises the following steps:
step 7-1: setting the initial value of the random sampling times a to be 0;
step 7-2: setting the random sampling times a as a + 1;
at t for household appliancesiRandomly sampling the a th time of the running state of the moment to obtain the dq of the j th type of household appliancejAt tiCoefficient of random sampling operation state at the a-th time
Random number if random sampling of a-th time
Then
Random number if random sampling of a-th time
Then
Is tiThe usage probability of the household appliance at any moment;
and 7-3: calculate user at tiThe load value of the a-th random sample of the time
And 7-4: if the random sampling frequency a is less than RS, returning to the step 1-2; if the random sampling frequency a is more than or equal to RS, calculating the load value p under the RS random samplingi；
The above-mentioned
RS is the maximum number of samples.
Compared with the closest prior art, the excellent effects of the invention are as follows:
the user load prediction method based on the electric quantity information provided by the invention utilizes the electric quantity collected by the intelligent electric meter at the peak load time every day, and optimizes the use probability of various household appliances by using a genetic algorithm; monte Carlo non-sequential random sampling was performed to simulate load distribution. The method can conveniently and quickly simulate and predict the load change of the user, improves the prediction accuracy, lays a foundation for reasonable configuration of the capacity of the distribution transformer of the distribution area, and improves the economic operation level and the construction efficiency of the distribution area.
Drawings
The invention is further described below with reference to the accompanying drawings.
FIG. 1: the user load prediction method in the embodiment of the invention adopts an application diagram;
FIG. 2: a flow chart of a genetic algorithm in an embodiment of the invention;
FIG. 3: the embodiment of the invention provides a flow chart of a Monte Carlo non-sequential random sampling method.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are illustrative and intended to be illustrative of the invention and are not to be construed as limiting the invention.
The user load prediction method based on the electric quantity information can conveniently and quickly predict the load change of the user in the peak load period, improves the prediction accuracy, lays a foundation for reasonable allocation of the capacity of the distribution transformer in the distribution area, and improves the economic operation level and the construction efficiency of the distribution area.
First, as shown in fig. 1, the user load prediction method specifically includes the steps of:
one) obtaining a category set DQ, a power set PDQ, a quantity set NDQ and an initial typical value matrix PB of the use probability of the household appliances of the user0A set of times T for peak load periods and a set of electricity usage Q for user peak load periods.
1. Family appliance category set DQ ═ { DQ ═ DQ1,dq2,...,dqj,...,dqnN is the total number of types of household appliances, dqjIs the name of the j-th type of household appliance.
2. Power set PDQ of household appliance (PDQ)1,pdq2,...,pdqj,...,pdqn}，pdqjIs the average power of the j-th type of household appliance.
3. Number of home appliances NDQ ═ NDQ1,ndq2,...,ndqj,...,ndqn}，ndqjThe number of the j-th type home appliances.
4. The set of times T ═ T at peak load hours1,t2,...,ti,...,tm}，tmIs a time value of 24 system time, t is more than or equal to 1m≤24。
5. Initial typical value matrix of use probability of household appliances
Is tiTime j type household appliance dqjUsing the initial value of the probabilistic typical value.
6. Set Q ═ Q of electricity consumption in peak load hours of the user1,q2,...,qk,...,qd}，qkThe electricity consumption of the user in the peak load period on the kth day is d, and the total days for collecting the electricity consumption by the intelligent electric meter are d.
Two) the time number i of the high peak load time in one day is set to 1, and in the embodiment, 1 is not less than i and not more than m.
Three) will use the probability typical value matrix PBi-1Assignment to matrix of use probability typical values PBiIn (1).
PBi-1Is ti-1Probability typical value matrix, PB, for use by a domestic appliance at a timeiIs tiThe household appliances at the moment use a probabilistic typical value matrix.
Four) according to the typical value matrix PB of the current household appliance use probabilityiMonte Carlo non-sequential random sampling is carried out to calculate the load value p of each moment of the user in the peak load periodiConstructing a set P of load curve sequence valuesi＝{p1,p2,...,pi,...,pm}。
Five) optimized use of the probability typical value matrix PBi。
As shown in FIG. 2, t is optimized by genetic algorithm in this embodimentiTypical value matrix PB of use probabilities of time-of-day household appliancesiThe method comprises the following specific steps:
1. setting the initial value of the iteration times h as 0, the maximum value of the iteration times h as maxnum, and the number of the population scale as ZQnum.
2. According to the gene coding strategy, t is dividediTypical value matrix PB of use probabilities of time-of-day household appliancesiEach using the probability to perform corresponding chromosome coding to generate an initial population ZQ0。
3. And setting the iteration number h as h + 1.
4. The population ZQh-1Assigns data to a population ZQh。
ZQh-1For the h-1 th iteration generated population, ZQhThe population produced for the h iteration.
5. Setting population ZQhThe initial value of the middle chromosome marker l is 0.
6. And (4) setting a chromosome marker position l as l + 1.
7. Slave population ZQhTaking out the l-th chromosome, and performing inverse coding on the chromosome to obtain tiProbability of use of each type of domestic appliance at a time, e.g. tiProbability of use of the j-th type of household appliance at time
8. As shown in fig. 3, according to the current tiThe using probability of each type of household electrical appliance at the moment is randomly sampled by a Monte Carlo non-sequential random sampling method to obtain tiThe load value p of the household appliance of the user at any momenti。
9. T obtained according to step 8iThe load value p of the household appliance of the user at any momentiUpdating the current load curve sequence value set Pi＝{p1,p2,...,pi,...,pm}。
10. According to the current tiSet of load curve sequence values P of timei＝{p1,p2,...,pi,...,pmCalculating the average value Pav of the load of the household appliances of the user in the high peak load period in the day,
11. calculating the average load set PG (PG) of the peak load time period of the user according to the electricity consumption set Q1,pg2,...,pgk,...,pgd}。
Wherein, pgkIs the average load value of the user on day k, pgk＝qkAnd m and d are the total days for the intelligent electric meter to collect the electricity consumption of the user.
12. The fitness function value epsilon of the l-th chromosome is calculated,
13. if the chromosome marker bit l is less than ZQnum, returning to the step 6; if the chromosome marker bit l is more than or equal to ZQnum, executing the step 14;
14. and (4) selecting.
Selecting ZQnum chromosomes with smaller fitness function values epsilon according to the fitness function values epsilon of all chromosomes in the population; in this embodiment, the chromosomes are arranged from small to large according to the numerical value of the fitness function value epsilon, and the steps 15 are executed after the 1 st to the ZQnum chromosomes are selected;
15. for population ZQhAnd carrying out the whole cross operation of the parent and the gemini single-point genes.
16. And (5) carrying out mutation.
Wherein, the mutation is to control whether the chromosome is mutated or not according to the mutation rate, and when the mutation is required, the gene which is required to be mutated is randomly selected. After determining the genes needing mutation, mutation should be carried out according to the dependence and mutual exclusion relationship of the genes.
17. According to the result of variationChromosome renewal population ZQh。
18. If the iteration times h are less than maxnum, returning to the step 3; if the iteration times h is larger than or equal to maxnum, the chromosome with the optimal fitness function value epsilon is subjected to inverse coding to obtain tiAnd optimizing the value of the use probability of the household appliance at the moment.
19. According to the above-mentioned tiThe optimized value of the use probability of the household appliance at the moment, and the typical value matrix PB of the use probability of the optimized household appliance is determinedi。
Six) the time index i of the peak load period is set to i + 1.
Seventhly) if i is less than or equal to m, returning to the step three), using the probability typical value matrix PBi-1Assignment to matrix of use probability typical values PBiPerforming the following steps; if i is larger than m, according to the optimized use probability typical value matrix PBiAnd calculating a predicted user load value based on the number of various types of household appliances owned by the user.
Second, as shown in fig. 3, in the embodiment, steps one), five) and seven), the user's t is calculated by using the monte carlo non-sequential random sampling methodiLoad value p of the domestic appliance at any timeiThe method comprises the following specific steps:
1. the initial value of the random sampling frequency a is set to 0.
2. The random sampling number a is set to a + 1.
At t for household appliancesiRandomly sampling the a th time of the running state of the moment to obtain the dq of the j th type of household appliancejAt tiCoefficient of random sampling operation state at the a-th time
Random number if random sampling of a-th time
Then
The household appliance is not operated; random number if random sampling of a-th time
Then
The household appliance operates;
is tiThe probability of use of the household appliance at the moment.
3. Calculate user at tiThe load value of the a-th random sample of the time
4. If the random sampling frequency a is less than RS, returning to the step 2; if the random sampling times a is more than or equal to RS, calculating the household appliance load value p of the user under the RS random sampling timesi；
RS is the maximum number of samples.
Thirdly, the specific process of load prediction for the peak load period in summer of the user by adopting the load prediction method provided by the invention is as follows:
one) determining a category set DQ and a power set P of the household applianceDQNumber set NDQThe initial typical value matrix of the use probability, the set of moments T of the peak load period of the user and the set of electric quantity used by the user Q.
1. Set of categories of household appliances:
DQ＝{dq1,dq2,dq3,dq4,dq5,dq6,dq7,dq8,dq9,dq10,dq11,dq12,dq13,dq14,dq15,dq16}：
dq1indicating electric lamp lighting devices, dq2Indicating induction cooker, dq3Indicating electric rice cooker, dq4Electric kettle and dq display5Denotes a microwave oven, dq6Indicating a range hood, dq7Indicating high-grade electric cooking utensils (including high-grade equipment such as electric baking pan), dq8Indicating electric fan, dq9Indicating air conditioning, dq10Indicating electric warmer, dq11Representing a television, dq12Indicating refrigerator, dq13Showing washing machine, dq14Indicating electric water heater, dq15Representing home computers, dq16An electric bicycle is shown.
2. Power set PDQ of household appliance (PDQ)1,pdq2,...,pdq16}。
3. Number of home appliances NDQ ═ NDQ1,ndq2,...,ndq16}。
4. The set of times T for peak load hours is {9,10,11,17,18,19,20,21 }.
5. Initial typical value matrix of use probability of household appliances
6. User electricity consumption set Q ═ Q1,q2,...,q62In this embodiment, the electricity consumption of the user in 7 and 8 months collected by the smart meter is selected.
Two) the time number i of the high peak load time in one day is set to 1, and in the embodiment, 1 is not less than i and not more than 8.
Three) will use the probability typical value matrix PBi-1Assignment to matrix of use probability typical values PBiIn (1).
Four) according to the typical value matrix PB of the current household appliance use probabilityiMonte Carlo non-sequential random sampling is carried out to calculate the load value p of each moment of the user in the peak load periodiForming a current load curve sequence value set Pi＝{p1,p2,...,pi,...,pm}。
Fifthly) power consumption Q ═ Q { Q } according to peak load period of the user every day in summer 7 months and 8 months1,q2,...,q62And the current load curve sequence value set Pi＝{p1,p2,...,pi,...,pmUsing genetic algorithm to step by step pair tiUsing probability typical value matrix PB for time pointiAnd (6) optimizing.
Six) updating the current home appliance use probability typical value matrix PB according to the optimized use probability valuei，
Seven) the time sequence number i of the peak load period is set to i + 1.
Eighthly), judging whether the time sequence number i of the peak load time period is larger than 8, and turning to the third step if the time sequence number i of the peak load time period is smaller than or equal to 8); if i is larger than 8, go to step nine).
Nine) randomly sampling the running states of all household appliances of the user at the moment according to the optimized household appliance use probability typical value matrix and the number of various types of household appliances owned by the user and Monte Carlo non-sequential random sampling, and predicting the load value of the user in summer peak period.
Finally, it should be noted that: the described embodiments are only some embodiments of the present application and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
Claims (2)
1. A user load prediction method based on electric quantity information is characterized by comprising the following steps:
step 1: obtaining a category set DQ, a power set PDQ, a quantity set NDQ and a use probability initial typical value matrix PB of the household appliances of the user0The peak load time set T and the user peak load time electricity consumption set Q collected by the intelligent ammeter;
step 2: setting the time sequence number i of the user in the peak load time period to be 1, wherein i is more than or equal to 1 and is less than or equal to m, and m is the total time of the peak load time period;
and step 3: will use a probability typical value matrix PBi-1Assignment to matrix of use probability typical values PBiPerforming the following steps; the PB isi-1Is ti-1Probability typical value matrix, PB, for use by a domestic appliance at a timeiIs tiThe household appliances at the moment use the probability typical value matrix;
and 4, step 4: according to the typical value matrix PB of the use probabilityiCalculating the load value of each moment in the peak load period of the user, and constructing a load curve sequence value set Pi＝{p1,p2,...,pi,...,pm}，piIs tiThe load value of the household appliance at the moment;
and 5: optimizing said usage probability typical value matrix PB by a genetic algorithmi；
Step 6: setting a time sequence number i ═ i +1 in a peak load period;
and 7: if i is less than or equal to m, returning to the step 3; if i is larger than m, the typical value matrix PB of the optimized household appliance use probability is usediCalculating and predicting the load value of the user by using a Monte Carlo non-sequential random sampling method according to the quantity of various types of household appliances owned by the user;
the set of kinds DQ of the household appliances in the step 1 is { DQ ═ DQ1,dq2,...,dqj,...,dqnN is the total number of types of household appliances, dqjName of the j type household appliance; the power set PDQ of the household appliance is { p ═ pdq1,pdq2,...,pdqj,...,pdqn}，pdqjThe average power of the j type household appliance; the number set NDQ of the household appliances is { NDQ ═1,ndq2,...,ndqj,...,ndqn}，ndqjThe number of the j type household appliances; the set of times of the peak load period T ═ T1,t2,...,ti,...,tm}，tmIs a time value of 24 system time, t is more than or equal to 1m≤24；
An initial typical value matrix of the usage probability of the household appliance
said step 5 is to optimize t by genetic algorithmiTypical value matrix PB of use probabilities of time-of-day household appliancesiThe method comprises the following steps:
step 5-1: setting the initial value of the iteration times h as 0, the maximum value of the iteration times h as maxnum, and the number of population scales as ZQnum; will tiTypical value matrix PB of use probabilities of time-of-day household appliancesiEach using the probability to perform corresponding chromosome coding to generate an initial population ZQ0；
Step 5-2: setting the iteration number h as h + 1;
step 5-3: the population ZQh-1Assigns data to a population ZQh；ZQh-1For the h-1 th iteration generated population, ZQhThe population generated for the h iteration; setting population ZQhThe initial value of the middle chromosome marker bit l is 0;
step 5-4: setting the chromosome marker bit l ═ l + 1;
step 5-5: the l chromosome is subjected to inverse coding to obtain tiThe use probability of each type of household appliance at any moment;
and 5-6: according to the said tiThe using probability of each type of household electrical appliance at the moment is obtained by a Monte Carlo non-sequential random sampling methodiLoad value p of the domestic appliance at any timeiAnd updating the load curve sequence value set Pi＝{p1,p2,...,pi,...,pm}；
And 5-7: according to the updated load curve sequence set Pi Calculating the load average value Pav of the household appliances of the user in the peak load period,
and 5-8: acquiring an average load set PG (PG) of peak load periods of a user according to the electricity consumption set Q1,pg2,...,pgk,...,pgd}；pgkD is the total number of days for collecting the electricity consumption of the user by the intelligent ammeter;
and 5-9: the fitness function value epsilon of the l-th chromosome is calculated,
step 5-10: if the chromosome marker bit l is less than ZQnum, returning to the step 5-4; if the chromosome marker bit l is more than or equal to ZQnum, executing the step 5-11;
step 5-11: arranging the chromosomes from small to large according to the numerical value of the fitness function value epsilon, selecting 1 st to ZQnum chromosomes and then executing the step 5-12;
and 5-12: for population ZQhCarrying out whole cross operation of the parent and the gemini single-point genes;
step 5-13: carrying out variation on chromosomes and updating the ZQ population according to the varied chromosomesh；
And 5-14: if the iteration times h are less than maxnum, returning to the step 5-2; if the iteration times h is larger than or equal to maxnum, the chromosome with the optimal fitness function value epsilon is subjected to inverse coding to obtain tiThe using probability optimization value of the household appliance at any moment;
in step 7, the method for calculating t by using Monte Carlo non-sequential random samplingiLoad value p of the domestic appliance at any timeiThe method comprises the following steps:
step 7-1: setting the initial value of the random sampling times a to be 0;
step 7-2: setting the random sampling times a as a + 1;
at t for household appliancesiRandomly sampling the a th time of the running state of the moment to obtain the dq of the j th type of household appliancejAt tiCoefficient of random sampling operation state at the a-th time
Random number if random sampling of a-th time
Then
Random number if random sampling of a-th time
Then
Is tiThe usage probability of the household appliance at any moment;
and 7-3: calculate user at tiThe load value of the a-th random sample of the time
And 7-4: if the random sampling frequency a is less than RS, returning to the step 7-2; if the random sampling frequency a is more than or equal to RS, calculating the load value p under the RS random samplingi；
The above-mentioned
RS is the maximum number of samples.
2. The method of claim 1, wherein the set of user power amounts Q ═ Q in step 11,q2,...,qk,...,qd}，qkD is the total number of days for collecting the electricity consumption by the intelligent ammeter;
in the step 4, the Monte Carlo non-sequential random sampling method is used for calculating tiOf appliances at all timesLoad value pi。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510335111.2A CN106326996B (en) | 2015-06-16 | 2015-06-16 | User load prediction method based on electric quantity information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510335111.2A CN106326996B (en) | 2015-06-16 | 2015-06-16 | User load prediction method based on electric quantity information |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106326996A CN106326996A (en) | 2017-01-11 |
| CN106326996B true CN106326996B (en) | 2021-09-03 |
Family
ID=57732129
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510335111.2A Active CN106326996B (en) | 2015-06-16 | 2015-06-16 | User load prediction method based on electric quantity information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106326996B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107292414B (en) * | 2017-05-02 | 2021-02-09 | 国网浙江省电力有限公司 | Resident user distributed internet optimal selection method based on peak-valley electric quantity prediction |
| CN107766992B (en) * | 2017-11-09 | 2021-07-20 | 上海电力学院 | Household daily load curve fine prediction method based on user behaviourology |
| CN110766186A (en) * | 2018-07-26 | 2020-02-07 | 珠海格力电器股份有限公司 | Method and device for predicting power consumption |
| CN109359780B (en) * | 2018-11-16 | 2021-10-08 | 上海电力学院 | Residential electricity consumption prediction method based on electrical appliance index |
| CN110110759B (en) * | 2019-04-15 | 2023-07-11 | 东南大学 | Multi-dimensional information identification-based power grid electrical information pointing method and system |
| CN112564085B (en) * | 2020-10-22 | 2023-04-18 | 国网山东省电力公司济宁供电公司 | Method and system for predicting maximum power load of electric heating distribution transformer |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103616862A (en) * | 2013-11-19 | 2014-03-05 | 青岛海尔软件有限公司 | Household appliance electricity consumption circumstance management system and method |
| CN104122819A (en) * | 2014-07-18 | 2014-10-29 | 东北电力大学 | User habit based intelligent household electricity utilization method and system thereof |
| CN104200297A (en) * | 2014-07-11 | 2014-12-10 | 浙江大学 | Energy optimizing dispatching method of home hybrid power supply system in real-time power price environment |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8103388B2 (en) * | 2009-01-29 | 2012-01-24 | International Business Machines Corporation | System for prediction and communication of environmentally induced power useage limitation |
-
2015
- 2015-06-16 CN CN201510335111.2A patent/CN106326996B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103616862A (en) * | 2013-11-19 | 2014-03-05 | 青岛海尔软件有限公司 | Household appliance electricity consumption circumstance management system and method |
| CN104200297A (en) * | 2014-07-11 | 2014-12-10 | 浙江大学 | Energy optimizing dispatching method of home hybrid power supply system in real-time power price environment |
| CN104122819A (en) * | 2014-07-18 | 2014-10-29 | 东北电力大学 | User habit based intelligent household electricity utilization method and system thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106326996A (en) | 2017-01-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106326996B (en) | User load prediction method based on electric quantity information | |
| Yu et al. | A systematic procedure to study the influence of occupant behavior on building energy consumption | |
| Aydinalp et al. | Modeling of the appliance, lighting, and space-cooling energy consumptions in the residential sector using neural networks | |
| Ghedamsi et al. | Modeling and forecasting energy consumption for residential buildings in Algeria using bottom-up approach | |
| Swan et al. | Occupant related household energy consumption in Canada: Estimation using a bottom-up neural-network technique | |
| Xue et al. | An interactive building power demand management strategy for facilitating smart grid optimization | |
| Zhang et al. | Response-surface-model-based system sizing for Nearly/Net zero energy buildings under uncertainty | |
| Yu et al. | A decision tree method for building energy demand modeling | |
| Ding et al. | Benchmark analysis of electricity consumption for complex campus buildings in China | |
| CN112561728B (en) | Comprehensive energy consumption optimization method, medium and device based on attention mechanism LSTM | |
| Guo et al. | Data mining based framework for exploring household electricity consumption patterns: A case study in China context | |
| CN104181898A (en) | Intelligent control method and system for interactive home appliances on basis of time-of-use electricity price response | |
| CN111598478B (en) | Comprehensive energy demand response quantity calculation method | |
| CN114418249A (en) | Operation control method and device for light storage flexible system | |
| CN107220735A (en) | A kind of multivariable rural power grids power predicating method of power industry classification | |
| JP2016177675A (en) | Estimation device, estimation method and computer program | |
| CN115081902A (en) | Comprehensive planning method, device, equipment and medium based on source network load-storage cooperation | |
| Luo et al. | Many-objective day-ahead optimal scheduling of residential flexible loads integrated with stochastic occupant behavior models | |
| Shen et al. | Exploring potential for residential energy saving in New York using developed lightweight prototypical building models based on survey data in the past decades | |
| Elnagar et al. | Framework to assess climate change impact on heating and cooling energy demands in building stock: A case study of Belgium in 2050 and 2100 | |
| CN105739308B (en) | Power optimization control method and system applied to temperature control electric appliance | |
| CN109559146B (en) | Provincial and municipal access data center optimization method for predicting number of potential users by electric power customer service center based on logistic model | |
| CN111859242A (en) | Household power energy efficiency optimization method and system | |
| Pedersen et al. | Handling thermal comfort in economic model predictive control schemes for demand response | |
| CN104380317A (en) | User management method, server device, and user management system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |